Highway vehicle detection method and device, electronic equipment and storage medium

Abstract

The application relates to an expressway vehicle detection method and device, electronic equipment and a storage medium, and belongs to the technical field of intelligent transportation. The application relates to an expressway vehicle detection method and device, electronic equipment and a storage medium, and belongs to the technical field of intelligent transportation. The application provides an expressway vehicle detection method, which can realize all-time and all-region monitoring of an expressway through a DAS system, realizes efficient detection of expressway vehicles, accurately and quickly locates a vehicle driving position and obtains a driving track.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to a method, apparatus, electronic device and storage medium for detecting vehicles on highways. Background Technology

[0002] With the increasing number of motor vehicles in my country, the demand for road transportation is constantly rising, putting enormous pressure on existing infrastructure and generating a series of traffic-related ecological problems. The huge traffic demand has resulted in a long-term supply shortage of road transportation in key areas, with road traffic volume operating close to maximum capacity for extended periods. Under these conditions, even minor disturbances can easily cause traffic ripples and congestion, significantly impacting the operational efficiency of the road traffic system. Therefore, real-time and reliable basic traffic vehicle information is fundamental for assessing traffic conditions and a prerequisite for achieving intelligent management of highways.

[0003] Existing vehicle detection technologies based on video, radar, or satellite are limited by factors such as weather and obstructions, as well as high costs. Summary of the Invention

[0004] In view of this, it is necessary to provide a method, device, electronic equipment and storage medium for detecting vehicles on highways, so as to achieve the purpose of accurately and quickly locating the vehicle's driving position and obtaining the driving trajectory.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for detecting vehicles on highways, comprising: Vibration signals of the highway traffic test section are acquired using a distributed fiber optic acoustic sensing system, and the time-domain and frequency-domain features of the vibration signals are extracted. A vehicle detection matrix is ​​constructed based on the time-domain features, the frequency-domain features, and the first preset threshold. The vehicle detection matrix is ​​input into the elastic network to obtain the vehicle driving trajectory map output by the elastic network. Based on the vehicle trajectory map, the vehicle detection results are obtained; Based on the vehicle detection results, proximity values ​​are used to determine the vehicle's trajectory information on the road segment to be tested.

[0006] In one possible implementation, extracting the time-domain features of the vibration signal includes: The vibration signal is subjected to Chebyshev filtering to obtain a filtered signal, and the filtered signal is then divided into frames to obtain a framed signal. The matched filtering result is obtained by performing correlation operations based on the predefined template signal and the framed signal, and the root mean square of the matched filtering result is calculated. The Cauchy distribution is fitted to the framed signal to obtain the Cauchy distribution location parameters and Cauchy distribution scale parameters; Calculate the variance of the framed signal; The time-domain features include the filtered signal, the root mean square, the Cauchy distribution location parameter, the Cauchy distribution scale parameter, and the variance.

[0007] In one possible implementation, extracting the frequency domain features of the vibration signal includes: Perform wavelet transform on the framed signal and solve for the wavelet coefficients; The power spectral density is calculated based on the autocorrelation function of the framed signal; The frequency domain features include the maximum value of the wavelet coefficients and the maximum value of the power spectral density.

[0008] In one possible implementation, constructing the vehicle detection matrix based on the time-domain features, the frequency-domain features, and the first preset threshold includes: When the root mean square is greater than the first preset threshold, it is determined that there is a vehicle in the test area corresponding to the root mean square, and it is marked as 2; When the root mean square is less than or equal to the first preset threshold, it is determined that there is no driving vehicle at the location of the test area corresponding to the root mean square, and it is marked as 1; The vehicle detection matrix is ​​obtained by combining the time-domain features, the frequency-domain features, and the labeling results.

[0009] In one possible implementation, obtaining the vehicle detection result based on the vehicle trajectory map includes: The vehicle region is determined based on the number of pixels in the vehicle trajectory map that are higher than the second preset threshold. Based on the pixels in the vehicle trajectory map that are below the second preset threshold, noise regions are determined; The data matrix values ​​of the vehicle region are marked as 2, and the data matrix values ​​of the noise region are marked as 1, to obtain a binary matrix; The vehicle detection result is obtained based on the binary matrix.

[0010] In one possible implementation, determining the vehicle's trajectory information on the road segment to be tested by calculating proximity values ​​based on the vehicle detection results includes: In the binary matrix, determine the location that is closest to and continuously marked as a vehicle area; The trajectory information is obtained based on the location of the nearest and consecutively marked vehicle area.

[0011] In one possible implementation, the wavelet coefficients are expressed as follows:

[0012] in, Represents wavelet coefficients, Indicates the framed signal. Describe the wavelet basis functions. The scale parameter represents the wavelet basis function. This represents the positional parameters of the wavelet basis functions.

[0013] Secondly, the present invention also provides a highway vehicle detection device, comprising: The acquisition unit is used to acquire vibration signals of the highway traffic test section through a distributed fiber optic acoustic wave sensing system, and extract the time-domain and frequency-domain features of the vibration signals. The construction unit is used to construct a vehicle detection matrix based on the time-domain features, the frequency-domain features, and a first preset threshold. The output unit is used to input the vehicle detection matrix into the elastic network to obtain the vehicle driving trajectory map output by the elastic network. The first detection unit is used to obtain vehicle detection results based on the vehicle driving trajectory map; The second detection unit is used to determine the vehicle's trajectory information on the road segment to be tested by performing proximity indexes based on the vehicle detection results.

[0014] Thirdly, the present invention also provides an electronic device, including a memory and a processor, wherein, The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the highway vehicle detection method described in any of the above implementations.

[0015] Fourthly, the present invention also provides a computer-readable storage medium for storing a computer-readable program or instructions, which, when executed by a processor, can implement the steps of the highway vehicle detection method described in any of the above implementations.

[0016] The beneficial effects of this invention are as follows: The highway vehicle detection method, device, electronic equipment, and storage medium provided by this invention acquire vibration signals of the highway traffic section to be tested through a DAS system, and calculate the time-domain and frequency-domain characteristics of the vibration signals. The DAS system enables real-time and full-domain monitoring of the highway. A vehicle detection matrix is ​​constructed based on the time-domain characteristics, frequency-domain characteristics, and a first preset threshold, and input into an elastic network to obtain a high-contrast vehicle trajectory map output by the elastic network. Furthermore, the vehicle area and noise area are distinguished based on the contrast of different areas in the vehicle trajectory map to obtain the vehicle detection result. Based on the vehicle detection result, the nearest neighbor index is used to accurately locate the vehicle, and complete trajectory information of each vehicle within the detection area is obtained, realizing efficient detection of highway vehicles, accurate and rapid positioning of vehicle driving position, and obtaining driving trajectory. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart of an embodiment of the highway vehicle detection method provided by the present invention; Figure 2 This is a basic schematic diagram of the DAS system used in this invention; Figure 3 A typical highway vibration signal waveform acquired by the DAS system provided by this invention; Figure 4 A typical comparison diagram of vehicle vibration signal and noise vibration signal provided by the present invention; Figure 5 A schematic diagram of the original signal provided by this invention after Chebyshev filtering; Figure 6 A schematic diagram of the matched filtering combined with root mean square result of the Chebyshev filtered signal provided by the present invention. Figure 7 This is a schematic diagram of the vehicle detection results provided by the present invention, which is based on setting a threshold using signal matched filtering combined with root mean square results. Figure 8 A schematic diagram comparing the histogram and fitting results of the framed signal provided by this invention; Figure 9 A schematic diagram comparing the Cauchy distribution fitting results of vehicle signal and noise signal after framing provided by the present invention; Figure 10A schematic diagram comparing the Gaussian distribution fitting results of vehicle signal and noise signal after framing provided by the present invention; Figure 11 A schematic diagram comparing the wavelet decomposition results of vehicle signal and noise signal after framing, as provided by the present invention; Figure 12 A schematic diagram comparing the power spectral density results of the autocorrelation function of the vehicle signal and noise signal after framing, as provided by this invention; Figure 13 A schematic diagram of the vehicle trajectory obtained by ridge-Lasso regression of the vibration signal and signal characteristics provided by the present invention. Figure 14 A schematic diagram illustrating the results of vehicle detection based on setting thresholds according to ridge-Lasso regression results provided by the present invention; Figure 15 A schematic diagram of the final vehicle inspection results provided by this invention; Figure 16 This is a schematic diagram of the structure of the highway vehicle detection system provided by the present invention; Figure 17 This is a schematic diagram of an embodiment of the highway vehicle detection device provided by the present invention; Figure 18 A schematic diagram of an embodiment of the electronic device provided by the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] In the description of the embodiments of the present invention, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0021] The terms "first," "second," etc., used in the embodiments of this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a technical feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] This invention provides a method, apparatus, electronic device, and storage medium for detecting vehicles on highways, which are described below.

[0024] Figure 1 This is a schematic flowchart of an embodiment of the highway vehicle detection method provided by the present invention, as shown below. Figure 1 As shown, the methods for detecting vehicles on highways include: S101. Obtain vibration signals of the highway traffic test section through a distributed fiber optic acoustic wave sensing system, and extract the time-domain and frequency-domain features of the vibration signals. S102. Construct a vehicle detection matrix based on the time-domain features, the frequency-domain features, and the first preset threshold; S103. Input the vehicle detection matrix into the elastic network to obtain the vehicle driving trajectory map output by the elastic network; S104. Based on the vehicle driving trajectory map, obtain the vehicle detection result; S105. Based on the vehicle detection results, perform proximity search values ​​to determine the vehicle's trajectory information on the road segment to be tested.

[0025] It should be noted that this invention utilizes a fiber optic distributed acoustic sensing (DAS) system to collect vibration signals from the entire highway section. Figure 2 This is a basic schematic diagram of the DAS system used in this invention, such as... Figure 2As shown, in the DAS system, a narrow-linewidth laser is used as the light source. The continuous light emitted by the narrow-linewidth laser is modulated into a fixed-pulse sequence by a photoelectric modulator. Then, the optical pulse sequence is amplified by an erbium-doped fiber amplifier (EDFA) and enters an ultra-weak fiber grating (UWFBG) array through the first circulator. The spacing between adjacent UWFBGs is 5m. The pulse sequence reflected from the UWFBG array enters an unbalanced Michelson interferometer through two circulators. In the Michelson interferometer, the optical path difference of the delay fiber and the spacing between adjacent UWFBGs are both 5m, used to compensate for the optical path difference between adjacent UWFBGs. The interferometer recovers the amplitude of the time-domain vehicle vibration signal by demodulating the phase change caused by the change in optical pulses between adjacent UWFBGs in the fiber. Finally, the three optical output signals of the 3×3 coupler are converted into electrical signals by three photodetectors for acquisition. The time-domain vehicle vibration signal is obtained through a photoelectric conversion process. In a DAS system, each pair of adjacent ultra-weak fiber gratings and the optical fiber between them constitutes a sensor, i.e., a monitoring area. In highway traffic, this system is used to collect vibration signals throughout the day within a road segment.

[0026] Figure 3 The following is a typical highway vibration signal waveform acquired by the DAS system provided by this invention. A typical vehicle vibration signal is shown below. Figure 3 As shown, this signal reflects vehicle information on highway traffic, such as the number of vehicles and the approximate travel time of vehicles in a certain measurement area.

[0027] Figure 4 This is a typical comparison diagram of vehicle vibration signals and noise vibration signals provided by the present invention. The comparison diagram shows typical vibration signals from vehicle movement and noise interference vibration signals monitored by the DAS system. Figure 4 As shown.

[0028] Compared to vibration signals with vehicles in motion, vibration signals with only noise interference have smaller fluctuations. However, since the four lanes of a highway are close together, they are inevitably affected by noise from vehicles traveling in adjacent lanes. This results in almost no difference in the fluctuation range between vibration signals with vehicles in motion and vibration signals with noise interference from adjacent lanes, making it difficult to detect vehicles based on the amplitude of the vibration signal. Therefore, it is essential to introduce features to characterize vehicle vibration signals.

[0029] In S101, the fiber Bragg grating used in DAS technology, due to its wavelength modulation characteristics, has sensing and detection capabilities that are not affected by fluctuations in the power of the light source of the sensing system or the bending loss of the optical fiber in the sensing network. Therefore, it can not only monitor the road surface in real time, 24 / 7, across the entire road section, and realize the perception and intelligence of vehicle operation status and road surface status of the entire road section, but also identify the status of highways.

[0030] For example, during the construction of a highway, two types of gratings, namely grating array vibration sensing optical cables and grating array strain sensing optical cables, can be buried along the construction site. Each grating measurement area can sense the vibration signal of the corresponding measurement area.

[0031] This invention utilizes a fiber optic distributed acoustic sensing system to collect vibration signals from the entire highway section. This system has advantages such as high sensitivity, distributed monitoring, strong anti-electromagnetic interference capability, low cost, and ease of large-scale reuse, enabling real-time and full-area monitoring of highways.

[0032] Furthermore, the time-domain and frequency-domain characteristics of the vehicle vibration signal are calculated. The time-domain characteristics of the signal are extracted by combining Chebyshev filtering, matched filtering, and the position parameters, scale parameters, and variance fitted by the root mean square and Cauchy distribution. The frequency-domain characteristics of the signal are extracted by wavelet transform and the power spectral density of the autocorrelation function.

[0033] In S102, since the signals sensed in a single vehicle vibration signal measurement area are random, it is necessary to use a large number of measurement areas to detect vehicles in order to obtain more accurate detection results.

[0034] In each test area, when the feature value corresponding to a certain location is greater than the first preset threshold, the location in that test area is considered to have a vehicle in motion and is marked as 2; otherwise, it is marked as 1, thus obtaining the marking result based on the threshold set by the feature value.

[0035] The vehicle detection matrix is ​​constructed by combining the vehicle signal feature values ​​(time domain features and frequency domain features) from multiple test areas with the labeling results.

[0036] In S103, Elastic Net is a regularization method for linear regression models that combines the advantages of L1 regularization (Lasso regression) and L2 regularization (ridge regression). It can handle multicollinearity and perform feature selection. The vehicle detection matrix is ​​input into the Elastic Net to obtain a high-contrast highway vehicle trajectory map output by the Elastic Net.

[0037] For example, the vehicle's trajectory is represented as a bright, sloping strip on a waterfall plot. This strip can be viewed as a ridge rising on a plane (or plain). The vehicle's trajectory can be effectively extracted by operating on this ridge using a ridge regression model.

[0038] However, the measured data inevitably contains impulse noise, which appears as discrete bright spots near the bright, sloping bands in the waterfall plot. If the ridge regression model is used directly, these discrete bright spots will significantly affect the fitting performance of the ridge regression.

[0039] The elastic network can effectively suppress the influence of discrete bright spots by introducing the Lasso model, thereby improving the fitting performance of vehicle trajectory and obtaining a high-contrast highway vehicle trajectory map.

[0040] In S104, an appropriate threshold is selected based on the different contrasts in different areas of the vehicle trajectory map. Pixels in the image above the threshold are marked as vehicle areas, and the data matrix at that location is marked as 2. Pixels below the threshold are marked as noise areas, and the data matrix value at that location is marked as 1. Based on the above process, the final detection result of the vehicle is obtained.

[0041] In step S105, proximity values ​​are used to accurately locate vehicles based on the vehicle detection results. The nearest and consecutive locations marked as vehicle areas are found in the binary matrix of the vehicle detection results and recorded. This process is repeated to obtain the complete trajectory information of each vehicle within the detection area.

[0042] In summary, the highway vehicle detection method provided by this invention acquires vibration signals from the highway traffic section under test through a DAS system, and calculates the time-domain and frequency-domain features of the vibration signals. The DAS system enables real-time, full-domain monitoring of the highway. A vehicle detection matrix is ​​constructed based on the time-domain features, frequency-domain features, and a first preset threshold, and then input into an elastic network to obtain a high-contrast vehicle trajectory map output by the elastic network. Furthermore, the contrast of different areas in the vehicle trajectory map is used to distinguish vehicle areas from noise areas to obtain vehicle detection results. Based on the vehicle detection results, proximity search is performed to accurately locate the vehicle, obtaining complete trajectory information for each vehicle within the detection area. This achieves efficient detection of highway vehicles, accurately and quickly locating the vehicle's position and obtaining its trajectory.

[0043] In some embodiments of the present invention, extracting the time-domain features of the vibration signal includes: The vibration signal is subjected to Chebyshev filtering to obtain a filtered signal, and the filtered signal is then divided into frames to obtain a framed signal. The matched filtering result is obtained by performing correlation operations based on the predefined template signal and the framed signal, and the root mean square of the matched filtering result is calculated. The Cauchy distribution is fitted to the framed signal to obtain the Cauchy distribution location parameters and Cauchy distribution scale parameters; Calculate the variance of the framed signal; The time-domain features include the filtered signal, the root mean square, the Cauchy distribution location parameter, the Cauchy distribution scale parameter, and the variance.

[0044] Specifically, the time-domain features include: obtaining a signal with less noise interference by performing Chebyshev filtering on the matrix composed of vibration signals; calculating the root mean square (RMS) of the signal after framing it and performing matching operations with a predefined signal; calculating the position and scale parameters of the Cauchy distribution by fitting the framing vibration signals to the Cauchy distribution; and calculating the variance of the matrix composed of the framing vibration signals.

[0045] For example, the steps for detecting vehicle vibration signals using time-domain features are as follows: First, regarding vibration signals Chebyshev filtering is performed to reduce the impact of noise interference and ensure signal quality. Figure 5 This is a schematic diagram of the original signal provided by the present invention after Chebyshev filtering. The signal after Chebyshev filtering is as follows: Figure 5 As shown, in Figure 5 In the diagram, the positions of the five vehicles are marked with C1, C2, C3, C4, and C5, respectively, while N1 and N2 mark the positions most affected by noise.

[0046] Due to significant interference from vehicles in adjacent lanes, the quality of the filtered signal remains poor. Therefore, it is necessary to improve signal performance by combining matched filtering with root mean square filtering. First, the signal is framed. With predefined template signals Perform correlation calculations to obtain the matched filtering results. as follows:

[0047] in N It is a signal The length of the signal is determined to enable signal extraction.

[0048] Then the result is obtained using matched filtering. Perform root mean square (RMS) calculation to obtain the signal energy information. To optimize signal quality, the calculation formula is as follows:

[0049] Figure 6 This is a schematic diagram of the matched filtering combined with root mean square (RMS) result of the Chebyshev filtered signal provided by the present invention. The signal obtained after matched filtering and RMS is as follows: Figure 6 As shown.

[0050] according to Figure 6 The results shown indicate that setting a threshold enables vehicle signal detection. Figure 7 This is a schematic diagram illustrating the vehicle detection results provided by the present invention, which uses a threshold set based on signal matched filtering combined with root mean square results. The detection results are as follows: Figure 7 As shown.

[0051] Because in some test areas the difference between the root mean square results of matched filtering for vehicles and noise is small, the following problems occur when setting the threshold using a single feature: lowering the threshold leads to increased noise interference, while increasing the threshold leads to missed vehicle detection. Therefore, it is necessary to select more feature values ​​to jointly characterize the signal in order to improve detection performance.

[0052] Second, the collected vibration signals are combined into a data matrix according to the test area, and the data matrix is ​​first divided into frames to obtain the matrix. Then, statistical analysis is performed to find the probability distribution model that best describes the data distribution. Visual histograms are drawn to understand the distribution characteristics of the data, and the fitting results of three common distributions—normal, Gaussian, and Laplace—are compared. Figure 8 This is a schematic diagram comparing the histogram and fitting results of the framed signal provided by the present invention. The fitting results are as follows: Figure 8 As shown.

[0053] Based on the fitting results, the Laplace distribution has sharp peaks that are inconsistent with the actual data distribution, while the fitting results of the Cauchy distribution and Gaussian distribution show smooth characteristics and are consistent with the overall trend of the data. Therefore, the relevant parameters of the Cauchy distribution and Gaussian distribution are selected to represent the vibration signal.

[0054] Furthermore, the Cauchy distribution location parameters, Cauchy distribution scale parameters, and Gaussian distribution variance parameters of the vibration signals from vehicles and the vibration signals from noise interference from adjacent lanes all differ, as shown in the following figures. Figure 9 and Figure 10 As shown, Figure 9 This is a schematic diagram comparing the Cauchy distribution fitting results of vehicle signals and noise signals after framing, as provided by the present invention. Figure 10 This is a schematic diagram comparing the Gaussian distribution fitting results of vehicle signals and noise signals after framing, as provided by the present invention.

[0055] In some embodiments of the present invention, extracting the frequency domain features of the vibration signal includes: Perform wavelet transform on the framed signal and solve for the wavelet coefficients; The power spectral density is calculated based on the autocorrelation function of the framed signal; The frequency domain features include the maximum value of the wavelet coefficients and the maximum value of the power spectral density.

[0056] In some embodiments of the present invention, the wavelet coefficients are expressed as follows:

[0057] in, Represents wavelet coefficients, Indicates the framed signal. Describe the wavelet basis functions. The scale parameter represents the wavelet basis function. This represents the positional parameters of the wavelet basis functions.

[0058] Specifically, the frequency domain features include: using wavelet transform to decompose the framed vibration signal into sub-signals of different frequencies and solving for the wavelet coefficients; calculating the correlation of the signal by constructing a linear matrix of the framed vibration signal, and further calculating the power spectral density of the autocorrelation function.

[0059] For example, the steps for detecting vehicle vibration signals using frequency domain features are as follows: First, select commonly used Daubechies wavelet functions for the data matrix. X Wavelet transform is performed to obtain wavelet decomposition coefficients, thereby effectively extracting information from the signal. The calculation formula is as follows:

[0060] in, These are the coefficients of the discrete wavelet transform. For wavelet basis functions, The scaling parameter determines the frequency of the wavelet basis functions. The position parameter determines the translation position of the wavelet basis function.

[0061] Wavelet transform decomposes a signal into wavelet components at different scales and locations, thereby enabling signal analysis and feature extraction. The wavelet transform results for vibration signals from vehicle movement and vibration signals from noise interference from adjacent lanes are shown below. Figure 11 As shown, Figure 11 This is a schematic diagram comparing the wavelet decomposition results of vehicle signals and noise signals after framing, as provided by the present invention.

[0062] Second, compared to a simple Fourier transform, the power spectral density of the autocorrelation function provides higher spectral resolution, enabling a more accurate description of the energy distribution of a signal at different frequencies. This allows for a more refined analysis of the signal's frequency domain characteristics. The calculation formula is as follows:

[0063] in, The length of the signal is given. The peak frequency is obtained by calculating the power spectral density of the autocorrelation function to characterize the vibration signal. The power spectral density results of the autocorrelation function of vibration signals from vehicle travel and vibration signals from noise interference from adjacent lanes are shown below. Figure 12 As shown, Figure 12This is a schematic diagram comparing the power spectral density results of the autocorrelation function of the vehicle signal and noise signal after framing, as provided by the present invention.

[0064] In some embodiments of the present invention, constructing a vehicle detection matrix based on the time-domain features, the frequency-domain features, and a first preset threshold includes: When the root mean square is greater than the first preset threshold, it is determined that there is a vehicle in the test area corresponding to the root mean square, and it is marked as 2; When the root mean square is less than or equal to the first preset threshold, it is determined that there is no driving vehicle at the location of the test area corresponding to the root mean square, and it is marked as 1; The vehicle detection matrix is ​​obtained by combining the time-domain features, the frequency-domain features, and the labeling results.

[0065] This embodiment combines multiple features and ridge-Lasso regression. First, the features obtained from the feature selection and extraction module—including the Chebyshev filtered signal, the matched filter combined with the root mean square, the location and scale parameters fitted by the Cauchy distribution, the variance, the maximum value of the wavelet transform coefficients, and the maximum power spectral density of the autocorrelation function—are combined with the RMS threshold-based classification results to form a data matrix. This data matrix is ​​then input into the elastic network as the final dataset, resulting in a high-contrast vehicle trajectory map with a clear visual representation of the vehicle region. The result is as follows: Figure 13 As shown, Figure 13 This is a schematic diagram of the vehicle trajectory obtained by ridge-Lasso regression of the vibration signal and signal feature quantities provided by the present invention.

[0066] In some embodiments of the present invention, obtaining the vehicle detection result based on the vehicle driving trajectory map includes: The vehicle region is determined based on the number of pixels in the vehicle trajectory map that are higher than the second preset threshold. Based on the pixels in the vehicle trajectory map that are below the second preset threshold, noise regions are determined; The data matrix values ​​of the vehicle region are marked as 2, and the data matrix values ​​of the noise region are marked as 1, to obtain a binary matrix; The vehicle detection result is obtained based on the binary matrix.

[0067] To address the issue of varying contrast in different areas of the vehicle's trajectory map, an appropriate threshold (i.e., the second preset threshold) is selected.

[0068] Pixels in the image that are above a second preset threshold are marked as vehicle regions, and the data matrix value at that location is marked as 2.

[0069] Pixels in the image that are below a second preset threshold are marked as noise regions, and the data matrix value at that location is marked as 1.

[0070] The vehicle inspection results were obtained based on the above process, such as Figure 14 As shown, Figure 14 This diagram illustrates the vehicle detection results provided by the present invention, which uses a threshold set based on ridge-Lasso regression results. It compares the detection results obtained using only matched filtering combined with root mean square (RMS) analysis. Figure 7 In comparison, the number of areas where vehicles were missed during inspection has been significantly reduced.

[0071] In some embodiments of the present invention, determining the vehicle's trajectory information on the road segment to be tested by performing proximity index calculations based on the vehicle detection results includes: In the binary matrix, determine the location that is closest to and continuously marked as a vehicle area; The trajectory information is obtained based on the location of the nearest and consecutively marked vehicle area.

[0072] Based on the vehicle detection results, each vehicle is precisely located. The vehicle detection result is a binary matrix, and the nearest neighbor search process involves finding the nearest and consecutive value of 2 that marks the vehicle region in the binary matrix.

[0073] After detecting the first vehicle entering the monitoring area, a proximity tracing process begins. This involves finding and recording the nearest vehicle marker area in the next monitoring area to the one in the previous area. Repeating this process yields the complete trajectory information for each vehicle within the detection area. The results are as follows: Figure 15 As shown, Figure 15 This is a schematic diagram of the final vehicle inspection result provided by the present invention, which can basically avoid the situation of vehicles being missed during inspection.

[0074] This invention utilizes a fiber optic distributed acoustic sensing system to collect data from the entire highway section. This system boasts advantages such as high sensitivity, distributed monitoring, strong resistance to electromagnetic interference, low cost, and ease of large-scale reuse, enabling real-time, all-area monitoring of highways. This invention achieves efficient vehicle detection on highways, improving vehicle detection performance while reducing the operation and maintenance costs of the highway traffic system.

[0075] The present invention also provides a highway vehicle detection system. Figure 16 This is a schematic diagram of the structure of the highway vehicle detection system provided by the present invention, as shown below. Figure 16 As shown, the highway vehicle detection system includes a vibration signal acquisition module, a feature selection and extraction module, a vehicle detection module, and a proximity value positioning module.

[0076] The vibration signal acquisition module is used to collect vibration signals from the entire highway traffic section. In the distributed acoustic wave sensing system, every two adjacent ultra-weak fiber Bragg grating arrays and the optical fiber between them form a sensor, which in turn builds a grating array sensing network to monitor the high-density vibration response along the entire highway lane. Each grating measurement area can sense the vibration signal of the corresponding measurement area.

[0077] The feature selection and extraction module is used to calculate the time-domain and frequency-domain features of the vibration signal.

[0078] The vehicle detection module is used for classifying highway traffic conditions. Since the signals sensed in a single vehicle vibration signal measurement area are random, a large number of measurement areas are needed for vehicle detection to obtain more accurate detection results.

[0079] The specific method is as follows: Multi-measurement area feature calculation. In each measurement area, when the feature value corresponding to a certain position is greater than a pre-set threshold, the measurement area and that position are considered to have a vehicle in motion and are marked as 2. Otherwise, they are marked as 1. The multi-measurement area vehicle signal features obtained by the feature selection and extraction module and the marking results obtained based on the feature values ​​are used as the final dataset. The dataset is input into an elastic network to obtain a high-contrast highway vehicle trajectory map. The threshold is set according to the brightness difference of different regions of the image. Based on the above process, a binary time-series detection map of vehicles with the measurement area as the horizontal axis, the travel time as the vertical axis, and the markings at the corresponding positions is obtained, and then the traffic state is classified.

[0080] The proximity positioning module is used to locate the specific position of a vehicle and obtain a simple vehicle trajectory after traffic condition classification.

[0081] First, determine and record the position of the vehicle detection result matrix with a value of 2 in the first test area. Then, based on the vehicle's driving direction, find the horizontal and vertical axis ranges of the neighboring test areas with a value of 2 and record them. This will give you the detection result of a complete vehicle. Then, find the second one, and so on, to get the detection result of each vehicle. This will allow you to locate the vehicle and obtain a simple trajectory.

[0082] The present invention also provides a method for detecting vehicles on highways, comprising the following steps: Step 1: Obtain vibration signals from the section of the highway to be tested; Step 2: Calculate the time-domain characteristic value of the vibration signal by combining time-domain characteristic analysis and calculate the frequency-domain characteristic value of the vibration signal by combining frequency-domain characteristic analysis; Step 3: Based on the matched filtering in the time domain features, the root mean square value is combined to set a threshold. The vehicle detection result matrix is ​​obtained according to the threshold. This matrix is ​​combined with the matrix of all the above feature values ​​as the final dataset matrix and input into the elastic network to obtain a high-contrast vehicle trajectory map. Step 4: Select an appropriate threshold for different regions in the vehicle trajectory map where the contrast is different. Mark the pixels in the image above the threshold as vehicle regions and mark the data matrix at that location as 2. Mark the pixels below the threshold as noise regions and mark the data matrix at that location as 1. Based on the above process, the final vehicle detection result is obtained. Step 5: Calculate proximity values ​​based on the vehicle detection results to accurately locate the vehicle. Find and record the nearest and consecutive locations marked as vehicle areas in the binary matrix of the vehicle detection results. Repeat this process to obtain the complete trajectory information of each vehicle within the detection area.

[0083] This invention first acquires road surface vibration signals using a DAS system. Then, it calculates the time-domain and frequency-domain characteristics of the acquired vibration signals, including the signal obtained after Chebyshev filtering, the RMS of the signal after matched filtering, the scale and location parameters fitted by the Cauchy distribution, the coefficients obtained from wavelet decomposition, the power spectral density of the signal's autocorrelation function, and the signal variance. Next, it calculates the time-domain and frequency-domain characteristics of vibration signals acquired from multiple measurement areas. Then, it obtains traffic state classification results using a binary time-series vehicle detection map. Finally, this invention uses proximity lookups on the traffic state classification results to locate vehicles traveling on highways.

[0084] To better implement the highway vehicle detection method in this embodiment of the invention, based on the highway vehicle detection method, correspondingly, as follows: Figure 17 As shown, this embodiment of the invention also provides a highway vehicle detection device, the highway vehicle detection device 1700 comprising: The acquisition unit 1701 is used to acquire the vibration signal of the highway traffic test section through a distributed fiber optic acoustic wave sensing system, and extract the time domain features and frequency domain features of the vibration signal. Construction unit 1702 is used to construct a vehicle detection matrix based on the time domain features, the frequency domain features and the first preset threshold; Output unit 1703 is used to input the vehicle detection matrix into the elastic network to obtain the vehicle driving trajectory map output by the elastic network; The first detection unit 1704 is used to obtain vehicle detection results based on the vehicle driving trajectory map; The second detection unit 1705 is used to determine the vehicle's trajectory information on the road segment to be tested by performing proximity values ​​based on the vehicle detection results.

[0085] The highway vehicle detection device 1700 provided in the above embodiments can realize the technical solutions described in the above highway vehicle detection method embodiments. The specific implementation principles of each module or unit can be found in the corresponding content of the above highway vehicle detection method embodiments, which will not be repeated here.

[0086] like Figure 18 As shown, the present invention also provides an electronic device 1800. The electronic device 1800 includes a processor 1801, a memory 1802, and a display 1803. Figure 18 Only some components of the electronic device 1800 are shown, but it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead.

[0087] In some embodiments, processor 1801 may be a central processing unit (CPU), microprocessor, or other data processing chip, used to run program code stored in memory 1802 or process data, such as the highway vehicle detection method of the present invention.

[0088] In some embodiments, processor 1801 may be a single server or a group of servers. The server group may be centralized or distributed. In some embodiments, processor 1801 may be local or remote. In some embodiments, processor 1801 may be implemented on a cloud platform. In some embodiments, the cloud platform may include a private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, internal cloud, multi-cloud, or any combination thereof.

[0089] In some embodiments, memory 1802 may be an internal storage unit of electronic device 1800, such as a hard disk or memory of electronic device 1800. In other embodiments, memory 1802 may also be an external storage device of electronic device 1800, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., provided on electronic device 1800.

[0090] Furthermore, the memory 1802 may include both internal storage units of the electronic device 1800 and external storage devices. The memory 1802 is used to store application software and various types of data installed on the electronic device 1800.

[0091] In some embodiments, display 1803 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an organic light-emitting diode (OLED) touchscreen. Display 1803 is used to display information from electronic device 1800 and to display a visual user interface. Components 1801-1803 of electronic device 1800 communicate with each other via a system bus.

[0092] In one embodiment, when processor 1801 executes the highway vehicle detection program in memory 1802, the following steps can be performed: Vibration signals of the highway traffic test section are acquired using a distributed fiber optic acoustic sensing system, and the time-domain and frequency-domain features of the vibration signals are extracted. A vehicle detection matrix is ​​constructed based on the time-domain features, the frequency-domain features, and the first preset threshold. The vehicle detection matrix is ​​input into the elastic network to obtain the vehicle driving trajectory map output by the elastic network. Based on the vehicle trajectory map, the vehicle detection results are obtained; Based on the vehicle detection results, proximity values ​​are used to determine the vehicle's trajectory information on the road segment to be tested.

[0093] It should be understood that when the processor 1801 executes the highway vehicle detection program in the memory 1802, in addition to the functions mentioned above, it can also perform other functions, as detailed in the description of the corresponding method embodiments above.

[0094] Furthermore, this embodiment of the invention does not specifically limit the type of electronic device 1800 mentioned. Electronic device 1800 can be a mobile phone, tablet computer, personal digital assistant (PDA), wearable device, laptop computer, or other portable electronic device. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable electronic device can also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the invention, electronic device 1800 may not be a portable electronic device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0095] Accordingly, embodiments of the present invention also provide a computer-readable storage medium for storing a computer-readable program or instruction. When the program or instruction is executed by a processor, it can implement the steps or functions of the highway vehicle detection method provided in the above-described method embodiments.

[0096] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0097] The above provides a detailed description of the highway vehicle detection method, apparatus, electronic device, and storage medium provided by the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for detecting vehicles on highways, characterized in that, include: Vibration signals of the highway traffic test section are acquired using a distributed fiber optic acoustic sensing system, and the time-domain and frequency-domain features of the vibration signals are extracted. A vehicle detection matrix is ​​constructed based on the time-domain features, the frequency-domain features, and the first preset threshold. The vehicle detection matrix is ​​input into the elastic network to obtain the vehicle driving trajectory map output by the elastic network. Based on the vehicle trajectory map, the vehicle detection results are obtained; Based on the vehicle detection results, proximity values ​​are used to determine the vehicle's trajectory information on the road segment to be tested.

2. The highway vehicle detection method according to claim 1, characterized in that, The extraction of the time-domain features of the vibration signal includes: The vibration signal is subjected to Chebyshev filtering to obtain a filtered signal, and the filtered signal is then divided into frames to obtain a framed signal. The matched filtering result is obtained by performing correlation operations based on the predefined template signal and the framed signal, and the root mean square of the matched filtering result is calculated. The Cauchy distribution is fitted to the framed signal to obtain the Cauchy distribution location parameters and Cauchy distribution scale parameters; Calculate the variance of the framed signal; The time-domain features include the filtered signal, the root mean square, the Cauchy distribution location parameter, the Cauchy distribution scale parameter, and the variance.

3. The highway vehicle detection method according to claim 2, characterized in that, The extraction of the frequency domain features of the vibration signal includes: Perform wavelet transform on the framed signal and solve for the wavelet coefficients; The power spectral density is calculated based on the autocorrelation function of the framed signal; The frequency domain features include the maximum value of the wavelet coefficients and the maximum value of the power spectral density.

4. The highway vehicle detection method according to claim 3, characterized in that, The process of constructing a vehicle detection matrix based on the time-domain features, the frequency-domain features, and the first preset threshold includes: When the root mean square is greater than the first preset threshold, it is determined that there is a vehicle in the test area corresponding to the root mean square, and it is marked as 2; When the root mean square is less than or equal to the first preset threshold, it is determined that there is no driving vehicle at the location of the test area corresponding to the root mean square, and it is marked as 1; The vehicle detection matrix is ​​obtained by combining the time-domain features, the frequency-domain features, and the labeling results.

5. The highway vehicle detection method according to claim 1, characterized in that, The process of obtaining vehicle detection results based on the vehicle trajectory map includes: The vehicle region is determined based on the number of pixels in the vehicle trajectory map that are higher than the second preset threshold. Based on the pixels in the vehicle trajectory map that are below the second preset threshold, noise regions are determined; The data matrix values ​​of the vehicle region are marked as 2, and the data matrix values ​​of the noise region are marked as 1, to obtain a binary matrix; The vehicle detection result is obtained based on the binary matrix.

6. The highway vehicle detection method according to claim 5, characterized in that, The step of determining the vehicle's trajectory information on the road segment to be tested by calculating proximity values ​​based on the vehicle detection results includes: In the binary matrix, determine the location that is closest to and continuously marked as a vehicle area; The trajectory information is obtained based on the location of the nearest and consecutively marked vehicle area.

7. The highway vehicle detection method according to claim 3, characterized in that, The expression for the wavelet coefficients is as follows: in, Represents wavelet coefficients, Indicates the framed signal. Describe the wavelet basis functions. The scale parameter represents the wavelet basis function. This represents the positional parameters of the wavelet basis functions.

8. A highway vehicle detection device, characterized in that, include: The acquisition unit is used to acquire vibration signals of the highway traffic test section through a distributed fiber optic acoustic wave sensing system, and extract the time-domain and frequency-domain features of the vibration signals. The construction unit is used to construct a vehicle detection matrix based on the time-domain features, the frequency-domain features, and a first preset threshold. The output unit is used to input the vehicle detection matrix into the elastic network to obtain the vehicle driving trajectory map output by the elastic network. The first detection unit is used to obtain vehicle detection results based on the vehicle driving trajectory map; The second detection unit is used to determine the vehicle's trajectory information on the road segment to be tested by performing proximity indexes based on the vehicle detection results.

9. An electronic device, characterized in that, Including memory and processor, among which, The memory is used to store programs; The processor, coupled to the memory, is configured to execute the program stored in the memory to implement the steps in the highway vehicle detection method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the steps in the highway vehicle detection method according to any one of claims 1 to 7.

Images